In the dynamic light scattering technique, the size distribution is derived from an exponential fitting of the measured correlogram. Typical DLS users however, tend to jump immediately to the distribution result, without looking first at the correlogram to insure that the quality of the raw data is acceptable. The obvious down side to this approach is that the fitting algorithms cannot always distinguish high from low quality raw data, and will generally give a size distribution result, even for raw data that indicates user attention is needed. When attempting to evaluate DLS data quality then, the best practice is to always look first at the measured correlogram.
Figure 1 shows a comparison of a high quality DLS correlogram to a few correlograms indicative of common light scattering problems. The high quality correlogram in this figure can be described as having a high amplitude (Y intercept) and a smooth exponential decay all the way to a single, flat, and zero baseline. The table below Figure 1 lists the symptoms, indications of, and the recommendations for each of the low quality correlograms shown in the figure.
Figure 1: Comparison of a high quality DLS correlogram to a few correlograms indicative of common light scattering problems.
While checking the measured correlogram should always be the first step when evaluating the quality of a DLS measurement, there are also a variety of analytical parameters that can be used. The "Intercept" and "Size Merit" are Zetasizer Nano software available parameters that can be used as quality metrics for the raw data. The Intercept is the amplitude or Y intercept of the intensity autocorrelation function, g1(t), and is calculated via the following expression, where g1(t2) and g1(t3) are the correlation values from the 2nd and 3rd channels of the digital correlator.
The Size Merit is representative of a normalized signal to noise ratio, expressed as a percentage. The Size Merit is calculated using the following expression, and should range between 0 and 100. Merit values < 50 indicated that the sample concentration may be too low; Merit values > 100 indicate the likely presence of number fluctuations or large contaminants.
The DTS software also includes parameters that can be used as metrics for the quality of the correlogram fitting algorithms. The "Cumulants Fit Error" and "Multimodal Fit Error" parameters are the c2 values for the fitting of the measured correlogram data using the Cumulants (single exponential) and the NNLS (multi-exponential) methods, with c2 being calculated according to the following expression.
The fit of the Cumulants and NNLS algorithms to the measured data can also be viewed graphically in the DTS software. This viewing is accomplished using a Residuals display and/or a Fit display. Examples of the Cumulants and NNLS residual and fit graphs for a DLS measurement of a 60 nm latex in DI water are given Figure 2. As evident in this example, both algorithms show low residuals, with the Cumulant fit being slightly better than the NNLS. This is in fact, what we would expect for a single size population sample of low polydispersity. For additional information on displaying these graphs, please check the Report Designer section of the Help file within the Zetasizer Nano software.
Figure 2: Example residual and fit displays for the Cumulant and NNLS methods for a DLS measurement of a 60 nm latex in DI water.
For additional questions or information regarding Malvern Instruments complete line of particle and materials characterization products, visit us at www.malvern.com.
